The problem is the investors. They love recurring revenue. Hence data mining apps and closed ecosystems with subscription fees. And bait and switch tactics like Hue where the conditions deteriorate over time (like needing an account now so they can track you better)
Companies don't really want to give us freedom of choice. They just want to take our money. Ideally through subscriptions so they have to do almost nothing for it but can cash in easily.
So all the "special sauce" doesn't have a technical reason but am economical one.
It's true that the trend in the industry definitely goes towards subscriptions and closed eco systems.
We at AirGradient see that very well that more and more companies force their users into the cloud and some monitors even refuse to show current air quality when not registered to the cloud. Imagine you would need to register your fever thermometer with an app in the cloud before you could measure your temperature...
Part of the problem is not only the investors in Matter but also VC funded smart home equipment manufacturers that are pressured to go down the recurring revenue business model.
At AirGradient [1] we want to demonstrate a counter example. Our fully certified air quality monitors are open source hardware and thus we provide the firmware code and they can be easily re-flashed with whatever software you like to use. With this openness a great community has emerged that maintains integrations to various systems like Home Assistant via ESPHome.
Thanks for trying a more customer-friendly approach!
I don't have an immediate need for one of these, but do you have any plans to expand to Europe or work with a local distributor? Inflated shipping cost & time, VAT & customs duties (+additional processing time) and a cumbersome return process if anything goes wrong, makes for a less than ideal shopping experience.
I literally just purchased your fully assembled kit a few days ago specifically because all the parts are open source and easy to hack. Right now I’m running Amazon air quality monitors with a homebridge plugin I had to patch myself because Amazon changed their API which broke everything.
Massive respect to that HA support. I’ve started running HA and it amazes me what is available on some cloud products once you remove the proprietary cloud from them. I used to be a Google home or bust, slowly I’ve been going the cloud agnostic route. Any Android device should work with Google home (for the wife) and both HomeKit + Home Assistant. I say HomeKit and Home Assistant because I really like Apple ease of use and the ability for home assistant to pipe other compatible sensors directly into HomeKit for example, ecobees sensors provide more sensor data via home assistant than they do in their regular app and in HomeKit. Via Home Assistant I can add this extra sensor data directly into HomeKit.
This is exactly the kind of product worth supporting. It's Christmas Day, so won't be buying today, but I've added this to the list of items for my Home Assistant.
I think I'd seen your product before on HN. It looks nice but like the others say, I'm in Europe so it's a no go for now. Will keep an eye on it though for when you decide to expand!
Looks perfect for me with one show stopper: no Home Kit support. I love the idea of Home Assistant but do not have the free time to pick up another service to support in our house.
Only if you buy another gateway that works with that other app. Like many enthusiasts do, yes. Like home assistant. But it's not an option for the average Joe.
In many cases non-recurring revenue is barely even acknowledged. It doesn’t matter. Recurring revenue is the only meaningful form from a valuation perspective.
I would counter this idea its all corporations fault with "the general public doesn't want to rotate a tire, let alone grow a potato."
We socially infantilized each other. Technology plays a huge role in that.
Voices of the past warned that extreme division of labor would result in humans that are lazier, stupider than the dumbest animal. People sit online screeching about the slightest social faux pas, do nothing about intentional destabilization of education, healthcare, and infrastructure by political parties.
We complain that every last mile problem, every edge case we encounter, was not solved already by Big Tech. Never occurs to us to solve them ourselves.
I grew up being taught how to rebuild cars from the frame up, fix houses for people in need struggling in Midwest winters, was mentored in physics, chemistry by college age aunts and uncles, had a grandfather who retired after decades of designing computers for IBM teach me analog circuit designs/implementations, and helped farmers birth/raise livestock, got a couple uni degrees a while back. I'm barely 40.
Meanwhile my office worker colleagues waste time and real resources driving for an oil change. They demand others solve their survival problems for them. Friends struggling in this economy have started chanting "If I'm just going to berated for not working 7 days/week for a rich person, may as well quit and work 7 days/week growing my own food, fixing my home."
Fingers crossed the fiat hallucination Boomers force us into serving is about to vanish with them. Physics does not put an obligation on the future to propagate their rhetoric.
I think people are, generally, smarter and more capable than ever.
Social media just highlights the dumdums.
IMO the real issue is lack of instilling a sense of agency at a young age.
You were taught how to build cars from a young age. I’m sure you were encouraged to try things yourself along the way. You learned that your actions have a direct impact on the world and that you have the ability to adjust your actions to adjust the impact.
(At least in the US) Education is so retention and information focused that we don’t teach _how_ to apply that information.
Because of this we grow up looking to others for how to do anything, even if we have the know how.
It’s usually not until later in life when people realize their actions can truly have an impact when applied correctly
The oil change thing always baffled the shit out of me.
I don’t drive, as it’s been largely both unnecessary and impractical with where I’ve lived in the past decade, but I know how to check and change the oil in a car, among other basic vehicle maintenance tasks. It’s not hard or time consuming.
> Fingers crossed the fiat hallucination Boomers force us into serving is about to vanish with them. Physics does not put an obligation on the future to propagate their rhetoric.
You could have used a relative educated in economics. Remember that just because you’re an expert in one field doesn’t make you one in another.
> If Matter can succeed in convincing companies to compete on the merits of their devices—not the incidental revenue that comes from being a single-system device provider, or phone app marketing...
...then in a few years, people won't even notice that they're simply scanning a code on their devices and adding them to whatever home control system they wish.
The sentence you (partially) quote starts with "if" and ends with "wish". Cough.
I just don't trust smart home stuff, it all seems like a big data collection nightmare at best, at worst you could have criminals and government nannies both spying on you 24/7. I still use incandescent lights though too.
I recently set up Home Assistant and a bunch of ESPHome-based sensor devices. It was pretty easy (for someone who's technical enough to write YAML, use SSH, and do some basic soldering) and now I have a "smart home" which is totally bullshit-free and cost like $80 all-in. It was a fun like weekend project I can recommend it!
This. I tried IKEA’s stuff and it drove me bonkers. We’ve had light switches for over 100 years. Why would I want to replace a perfectly good light switch with flaky buttons and smart bulbs that need to be repaired every month?
But I still wanted to use Home Assistant to control lights. The solution? Smart light _switches_ and wall plugs.
I went with Lutron Caseta wall switches and plug-in dimmers. (Expensive at $60/switch.) Lutron has been doing commercial automation for a very long time and it shows.
1. They work without any automation. Most dimmable LEDs will work.
2. The remotes connect directly to the switches and don’t need a hub.
3. The pro hub model connects to home assistant. You can use unpaired remotes as switches. (Using the correct blueprints.)
I’ve had this set up for over a year and I’ve yet to replace a battery on the many dozens of remotes I have. And I’ve never had a problem with the switch not working.
I get to do all the tinkering I want without breaking my house. :)
Note: I do have some Phillips Hue bulbs for color but those are in lamps separate from overhead lighting.
If I want a light to be on, I turn it on. I don't go much for mood lighting, I have lamps and overhead lights and depending on the level of illumination I want or need at the time, one, some, or all might be powered up. But I've never thought of getting up from my ass groove on the couch and going over and operating a switch is a huge deal. And I'm pretty lazy.
For sure. Don’t use it. But some of us find it enjoyable to write automation scripts and get our homes behaving exactly how we want. Might I suggest you refrain from shitting on some one else’s hobby just because it’s not for you? I’m sure building a ship in a bottle “wastes” lots of time too.
I really like being able to have different light moods throughout the day, in different lamps around the room, instead of a single lamp in the ceiling. This would be very tricky to do with dimmable switches.
That’s one of the things Matter is supposed to help with - part of the spec is that all devices need to have a local-only / no-cloud / no proprietary app option
An ESP32 (or 8266 honestly in most scenarios) and ESPHome/WLED/anything that supports MQTT will do the job reliably and without concerns of data collection/spying. Isolate their network and open a one way firewall port for the proxy/dashboard you want to use. For most devices, soldering is minimal and simple. You spend a little more time on initial setup, but all those concerns slip away.
The first time I was reading about matter, I was also watching the last season of The Expanse (a hard sci-fi series). In that show, technology mostly just happens in the background, despite being central to the plot often.
I think that is because the characters tend to have a very nonchalant way of using it: Whatever terminal is nearest can be used to interact with stuff in vicinity. There are no boundaries between devices, services, or the like; everything communicates with everything else, over any channel available, without further thinking.
If we ever get past showing off how the ceiling bulb can do red light, maybe we could start having a protocol for automatic device capability negotiation/discovery?
There is zero incentive to create or support such a thing. No recurring revenue or lock in opportunities. It could only happen if it were mandated.
Open source could do it in theory, but only for nerds. Open source can’t create mass adoptable products because making technology easy to use is fabulously expensive and difficult.
So if the market can't achieve this, regulate. We live in a world built on standards, but industries like IoT have developed since the US went batshit and decided that sort of thing amounts to impurifying precious bodily fluids. Maybe the EU will do something to sort it out.
Some things shouldn’t depend on market incentive. This is a fundamental problem with technology, which by and large is unaware of any other technology than itself. This makes it burdensome to use several devices together in a useful manner. Going forward, I don’t think anyone would dispute there’s going to be more tech around. Maybe it’s time for something truly progressive, and mandate this?
I don’t think the market alone would create a unified rail or telephone network. Yet that has brought immense benefits for society at large, and vendors still manage to capitalise on top of the plumbing.
The only way this would ever work is if we have a body with authority that defines/approves standards (I guess with industry input), and devices that claim to do X must be certified, and certification must be clear during purchase. Customers can then choose to only buy certified devices. I’m very willing to sacrifice some flashy feature for interoperability.
And this should be across domains. A data or diagnostics scanner for example should be able to retrieve data from my car, from my fridge and from my hear rate monitor.
if you have an account online somewhere you can log in from any device. the device itself is merely a machine to run the code downloaded from your online account. you could do everything though that account and treat the device like a dumb terminal.
any specific device that provides special functions, like your car, could also connect to that account and upload its data there, so you can access it from anywhere else.
AFAIK, the purposes of the web was sharing information, building a corpus of knowledge by everyone interacting with each other, and it was so for sometime until the advent of the algorithms and walled gardens. I started on the web a bit late, and it was because I wanted to learn how to draw and play the guitar. There was a lot of weird sites build on Wordpress sharing tutorials and multiple forums with people interacting just like this one. Soon after FB came and almost everyone switches to pages and groups, and some years after you can barely find anything on google.
I wouldn't mind connected services, but when my data is held hostage and the one in control is a faceless company, it does not feel like technology anymore, something that would help you accomplish your task better. Take my smart tv, I would be glad to connect it to my network and not bother with the apple tv when I want to play a movie, but the last time I tried, it tried to forcefully upgrade itself and wouldn't let me skip some onboarding screen that tried to get me to create an account. I had to unplug it. Or when some headphones or mouse manufacturers want me to use bloated proprietary apps just to configure them.
While what you describe is technically feasible, it never feel like this as tech companies always start a conflict with their (paying) users by taking their data hostage, enshitifying their platforms, and destroying any attempt of integration.
Yes, well, mostly: I’d expect you’d have to add new things to your network with some level of confirmation once–if anything, proper auth should be baked in.
But after that, wouldn’t it be great if you didn’t have separate devices anymore, but a single digital identity across all of them, with an interface to control unique capabilities of each of them?
For a banal example - there’s no reason we shouldn’t be able to set a timer on our phone, and check it on a computer, or kitchen counter display. On the other hand, I can’t check on the progress of my vacuum robot from my laptop, because the vendor only offers a mobile app. I can’t pull the stats from my car at all, I can’t check a hard drive without plugging it into something compatible, I can’t easily send a big file to someone nearby. Technology should allow us to do all that.
Devices should promote their capabilities and offer a standard interface to control them, and simultaneously be able to display controls for the interface of other devices, depending on available means of input. Done well, this can actually be zero config.
I understand how compelling the marketing hype seems at first, but a quick investigation reveals this to be yet another mafia shakedown and effort to lock down and prevent any meaningful competition.
They have a 16-bit Vendor ID. That's 65,535 vendors supported. Globally. Ever. One must pay the CSA ([0] $2,500 per PRODUCT or $7,000 per YEAR) for one of these artificially scarce identifiers.
In order to download specs, you're expected to accept TOS, authenticate and then you're emailed a (presumably authenticated) link to the PDF [1].
You need to wait for the consortium to define every single application layer detail for each and every type of thing [2].
Of course, the decision makers are unable to make high quality decisions since their own businesses depend on trapping customers in their respective walled gardens. The goal is to prevent any small or medium sized businesses from competing with the small handful of multinational megacorps.
From what I can tell, it's all of the worst parts of USB, without HID and Serial to fall back on for things they didn't think to define.
> They have a 16-bit Vendor ID. That's 65,535 vendors supported. Globally. Ever. One must pay the CSA ([0] $2,500 per PRODUCT or $7,000 per YEAR) for one of these artificially scarce identifiers.
I didn't know any of this. I pay attention to IoT (and run a modest Home Assistant setup in my home.) Had I known about these constraints, I would have written off Matter from the beginning.
Coming from software and making the move to hardware has been incredibly eye-opening.
As soon as you need to instantiate something with atoms, the cycle-times and overall difficulty increases exponentially. Just when you think you've got something, all these despicable gatekeepers shove their hands in your pockets.
There are lots of reasons we can't have small batch, niche electronic products, and these license fees combined with even worse regulatory costs require enormous sales volumes to amortize.
It's really concerning to see the same rent-seeking values making such strong headway in the software space over the past 5-10 years too.
I think that the vendor ID limitation is less of a hurdle, realistically, than it’s made out to be, especially in the short term. Simple cert costs and loss of captured markets/lack of demand are the bigger impediments. But I don’t think the obituary needs to be written on matter yet.
The problem is that such a small address space, combined with such large fees, makes it impossible for almost all individuals (and tiny organizations) to participate on equal footing with megacorps.
Another measly byte (or two) could have made it possible for fees to be reduced down the road, but as-is, there is an artificial barrier to popular access.
The USB-IF also artificially limits the vendor address space, which gives them the ability to charge exorbitant fees before anyone can deploy a device without conflicts.
I haven't tried to build and deploy a PCI device yet, but given your point, I assume the same game is being played.
At least with these 2 technologies, they can lean on being around for some decades. Bits were quite a lot more expensive 20+ years ago.
What an odd choice for them to hide the download link behind a verified email address and coerce a TOS accept that is presumably linked, then make the documents available via search links.
I assumed my link in my email was fingerprinted as there are some odd characters encoded into the URL.
As it turns out, it matched the search result link.
I was so excited about Matter at first, but now it seems like nobody really cares.
WiFi has gotten way more reliable, and devices with open local APIs are more common now. Bluetooth is even starting to be used occasionally. YoLink's LoRa based hardware is still better than anything else even slightly affordable, even though it's still totally a closed cloud based platform.
I am also invested into Z-Wave with about 50 devices. It works well, doesn't require an Internet connection, degrades gracefully if the hub is unavailable (or indeed any RF) and Z-Wave JS UI is good enough for self-hosting and bridging to MQTT for full integration with anything else I want.
The protocol is also excellent for battery powered devices in terms of super low power consumption: depending on the device's needs, the protocol allows for the device radio to be completely off for hours at a time. For example, I have a couple of key fob controllers that we use daily but we haven't replaced the batteries in them for years.
My only fear is that "consumerist" IoT platforms that change every few years will usurp it and new Z-Wave devices will become unavailable.
I am using z-wave as well. It has been rock solid, and using the z-wave association capability I don’t even need the hub (using Hubitat) unless I want to do more complex automations. The only downside is the ecosystem has fewer options (but have been high quality IMO) and the cost is higher than the ubiquitous WiFi smart home devices that require their own apps and internet connection.
FWIW, Gardena sells "dumb" irrigation valves that run on 24V. It's just a normally-closed solenoid valve IIUC, with proper fittings etc. Controlling these with a 24V power supply and a Z-wave relay should be easy?
I'm reserving my interest in smart home tech for when it's a clear win in terms of energy efficiency. Very recently I compared my current home lighting costs with a fully smart-equipped setup, and it basically just balanced out. I was surprised, because I calculated a best case scenario where the lights could basically always be off when not in use, compared to my current patterns where honestly I'm pretty lazy and leave all them on almost all day. The industry/general consumer priorities are pretty frustrating.
I guess you have plenty of lights that you don’t leave on “all the time” such as lights in toilets that typically run at low duty cycles (I guess the light in my toilet is on for less than ten minutes a day, and only in winter, for example)
For such lights, the permanent power usage of the controller can be much larger than that of the light, even if you accidentally leave it on way longer than needed now and then.
So, that’s about 10Wh a day. If you use that to control a 10W light, you’d have to decrease its ‘on’ time by a hour a day to come out positive in energy usage. I have many lights that aren’t even on that long each day.
I think home automation of lighting fixtures as a way to conserve energy made way more sense before LEDs became common.
Yeah, that's what I was factoring. For my home, the 24/7 power consumption from a handful of switches + some kind of hub just balanced out the potential savings from having the same bulbs off more often.
Readit News
Companies don't really want to give us freedom of choice. They just want to take our money. Ideally through subscriptions so they have to do almost nothing for it but can cash in easily.
So all the "special sauce" doesn't have a technical reason but am economical one.
We at AirGradient see that very well that more and more companies force their users into the cloud and some monitors even refuse to show current air quality when not registered to the cloud. Imagine you would need to register your fever thermometer with an app in the cloud before you could measure your temperature...
Part of the problem is not only the investors in Matter but also VC funded smart home equipment manufacturers that are pressured to go down the recurring revenue business model.
At AirGradient [1] we want to demonstrate a counter example. Our fully certified air quality monitors are open source hardware and thus we provide the firmware code and they can be easily re-flashed with whatever software you like to use. With this openness a great community has emerged that maintains integrations to various systems like Home Assistant via ESPHome.
[1] https://www.airgradient.com/
I don't have an immediate need for one of these, but do you have any plans to expand to Europe or work with a local distributor? Inflated shipping cost & time, VAT & customs duties (+additional processing time) and a cumbersome return process if anything goes wrong, makes for a less than ideal shopping experience.
This way you’d get a list of potential customers and we’d get a notification when we’d get an easier option to buy your products.
There isn't an HN promo code is there?
Deleted Comment
We socially infantilized each other. Technology plays a huge role in that.
Voices of the past warned that extreme division of labor would result in humans that are lazier, stupider than the dumbest animal. People sit online screeching about the slightest social faux pas, do nothing about intentional destabilization of education, healthcare, and infrastructure by political parties.
We complain that every last mile problem, every edge case we encounter, was not solved already by Big Tech. Never occurs to us to solve them ourselves.
I grew up being taught how to rebuild cars from the frame up, fix houses for people in need struggling in Midwest winters, was mentored in physics, chemistry by college age aunts and uncles, had a grandfather who retired after decades of designing computers for IBM teach me analog circuit designs/implementations, and helped farmers birth/raise livestock, got a couple uni degrees a while back. I'm barely 40.
Meanwhile my office worker colleagues waste time and real resources driving for an oil change. They demand others solve their survival problems for them. Friends struggling in this economy have started chanting "If I'm just going to berated for not working 7 days/week for a rich person, may as well quit and work 7 days/week growing my own food, fixing my home."
Fingers crossed the fiat hallucination Boomers force us into serving is about to vanish with them. Physics does not put an obligation on the future to propagate their rhetoric.
IMO the real issue is lack of instilling a sense of agency at a young age.
You were taught how to build cars from a young age. I’m sure you were encouraged to try things yourself along the way. You learned that your actions have a direct impact on the world and that you have the ability to adjust your actions to adjust the impact.
(At least in the US) Education is so retention and information focused that we don’t teach _how_ to apply that information.
Because of this we grow up looking to others for how to do anything, even if we have the know how.
It’s usually not until later in life when people realize their actions can truly have an impact when applied correctly
I don’t drive, as it’s been largely both unnecessary and impractical with where I’ve lived in the past decade, but I know how to check and change the oil in a car, among other basic vehicle maintenance tasks. It’s not hard or time consuming.
You could have used a relative educated in economics. Remember that just because you’re an expert in one field doesn’t make you one in another.
Dead Comment
> If Matter can succeed in convincing companies to compete on the merits of their devices—not the incidental revenue that comes from being a single-system device provider, or phone app marketing...
The sentence you (partially) quote starts with "if" and ends with "wish". Cough.
Convenience ? One minute of me debugging a http request to fix my light is evaluated dearly.
If that happen once it will offset years of voice light off/on and other lazy ass of poor design workaround.
But I still wanted to use Home Assistant to control lights. The solution? Smart light _switches_ and wall plugs.
I went with Lutron Caseta wall switches and plug-in dimmers. (Expensive at $60/switch.) Lutron has been doing commercial automation for a very long time and it shows.
1. They work without any automation. Most dimmable LEDs will work.
2. The remotes connect directly to the switches and don’t need a hub.
3. The pro hub model connects to home assistant. You can use unpaired remotes as switches. (Using the correct blueprints.)
I’ve had this set up for over a year and I’ve yet to replace a battery on the many dozens of remotes I have. And I’ve never had a problem with the switch not working.
I get to do all the tinkering I want without breaking my house. :)
Note: I do have some Phillips Hue bulbs for color but those are in lamps separate from overhead lighting.
Not just with HA, but tech in general really. I love the idea of FOSS, but it's important that stuff just works, every time, without hackery.
I think that is because the characters tend to have a very nonchalant way of using it: Whatever terminal is nearest can be used to interact with stuff in vicinity. There are no boundaries between devices, services, or the like; everything communicates with everything else, over any channel available, without further thinking.
If we ever get past showing off how the ceiling bulb can do red light, maybe we could start having a protocol for automatic device capability negotiation/discovery?
Open source could do it in theory, but only for nerds. Open source can’t create mass adoptable products because making technology easy to use is fabulously expensive and difficult.
I don’t think the market alone would create a unified rail or telephone network. Yet that has brought immense benefits for society at large, and vendors still manage to capitalise on top of the plumbing.
And this should be across domains. A data or diagnostics scanner for example should be able to retrieve data from my car, from my fridge and from my hear rate monitor.
if you have an account online somewhere you can log in from any device. the device itself is merely a machine to run the code downloaded from your online account. you could do everything though that account and treat the device like a dumb terminal.
any specific device that provides special functions, like your car, could also connect to that account and upload its data there, so you can access it from anywhere else.
technically this is already possible today.
I wouldn't mind connected services, but when my data is held hostage and the one in control is a faceless company, it does not feel like technology anymore, something that would help you accomplish your task better. Take my smart tv, I would be glad to connect it to my network and not bother with the apple tv when I want to play a movie, but the last time I tried, it tried to forcefully upgrade itself and wouldn't let me skip some onboarding screen that tried to get me to create an account. I had to unplug it. Or when some headphones or mouse manufacturers want me to use bloated proprietary apps just to configure them.
While what you describe is technically feasible, it never feel like this as tech companies always start a conflict with their (paying) users by taking their data hostage, enshitifying their platforms, and destroying any attempt of integration.
But after that, wouldn’t it be great if you didn’t have separate devices anymore, but a single digital identity across all of them, with an interface to control unique capabilities of each of them?
For a banal example - there’s no reason we shouldn’t be able to set a timer on our phone, and check it on a computer, or kitchen counter display. On the other hand, I can’t check on the progress of my vacuum robot from my laptop, because the vendor only offers a mobile app. I can’t pull the stats from my car at all, I can’t check a hard drive without plugging it into something compatible, I can’t easily send a big file to someone nearby. Technology should allow us to do all that.
Devices should promote their capabilities and offer a standard interface to control them, and simultaneously be able to display controls for the interface of other devices, depending on available means of input. Done well, this can actually be zero config.
I understand how compelling the marketing hype seems at first, but a quick investigation reveals this to be yet another mafia shakedown and effort to lock down and prevent any meaningful competition.
They have a 16-bit Vendor ID. That's 65,535 vendors supported. Globally. Ever. One must pay the CSA ([0] $2,500 per PRODUCT or $7,000 per YEAR) for one of these artificially scarce identifiers.
In order to download specs, you're expected to accept TOS, authenticate and then you're emailed a (presumably authenticated) link to the PDF [1].
You need to wait for the consortium to define every single application layer detail for each and every type of thing [2].
Of course, the decision makers are unable to make high quality decisions since their own businesses depend on trapping customers in their respective walled gardens. The goal is to prevent any small or medium sized businesses from competing with the small handful of multinational megacorps.
From what I can tell, it's all of the worst parts of USB, without HID and Serial to fall back on for things they didn't think to define.
[0] https://csa-iot.org/become-member/#Membership
[1] https://csa-iot.org/developer-resource/specifications-downlo...
[2] You must download the "Device Library" spec yourself, as my link is fingerprinted and I do not wish to violate a hidden TOS clause.
[edit] formatting
This is why I love HN, thanks for pointing it out.
I didn't know any of this. I pay attention to IoT (and run a modest Home Assistant setup in my home.) Had I known about these constraints, I would have written off Matter from the beginning.
What were they thinking?
As soon as you need to instantiate something with atoms, the cycle-times and overall difficulty increases exponentially. Just when you think you've got something, all these despicable gatekeepers shove their hands in your pockets.
There are lots of reasons we can't have small batch, niche electronic products, and these license fees combined with even worse regulatory costs require enormous sales volumes to amortize.
It's really concerning to see the same rent-seeking values making such strong headway in the software space over the past 5-10 years too.
Take a gue$$
Another measly byte (or two) could have made it possible for fees to be reduced down the road, but as-is, there is an artificial barrier to popular access.
Isn't that the same as USB or PCI? Both also have a 16-bit vendor ID.
The USB-IF also artificially limits the vendor address space, which gives them the ability to charge exorbitant fees before anyone can deploy a device without conflicts.
I haven't tried to build and deploy a PCI device yet, but given your point, I assume the same game is being played.
At least with these 2 technologies, they can lean on being around for some decades. Bits were quite a lot more expensive 20+ years ago.
I assumed my link in my email was fingerprinted as there are some odd characters encoded into the URL.
As it turns out, it matched the search result link.
That's good to know.
Thanks!
WiFi has gotten way more reliable, and devices with open local APIs are more common now. Bluetooth is even starting to be used occasionally. YoLink's LoRa based hardware is still better than anything else even slightly affordable, even though it's still totally a closed cloud based platform.
Deleted Comment
Zigbee is also pretty cool, between these and an odd Wi-Fi only device here and there, it seems more than adequate.
The protocol is also excellent for battery powered devices in terms of super low power consumption: depending on the device's needs, the protocol allows for the device radio to be completely off for hours at a time. For example, I have a couple of key fob controllers that we use daily but we haven't replaced the batteries in them for years.
My only fear is that "consumerist" IoT platforms that change every few years will usurp it and new Z-Wave devices will become unavailable.
I hit a problem where it was really difficult to debug setup / configuration problems.
I suspect this was due to zwave devices being as cheap as possible: my devices had just 1-2 buttons and 1-2 blinky lights for configuration.
I would have gladly spent $30 more per device to get better device diagnostics.
The Z-wave devices turned out to be significantly more unreliable surprisingly.
The only z-wave device I still have running is my smart lock. Everything else has been replaced with zigbee.
All wired Z-Wave devices also act as repeaters to extend the network, which is handy.
Also important: not overdoing it. I don’t have tons of time to tinker with devices, HA, etc.
For such lights, the permanent power usage of the controller can be much larger than that of the light, even if you accidentally leave it on way longer than needed now and then.
The EU limit allows each controller to use 0,5w. See https://commission.europa.eu/energy-climate-change-environme.... And they do use close to that. The Philips Hue bulb uses 0,4W when off. See https://www.anandtech.com/show/6805/philips-hue-automated-ho....
So, that’s about 10Wh a day. If you use that to control a 10W light, you’d have to decrease its ‘on’ time by a hour a day to come out positive in energy usage. I have many lights that aren’t even on that long each day.
I think home automation of lighting fixtures as a way to conserve energy made way more sense before LEDs became common.